Giardia lamblia is among the commonest intestinal protozoa world-wide and may cause significant morbidity, especially in children. Although 5-nitroimidazole compounds have formed the mainstay of giardiasis treatment for several years, the increasing number of reports of refractory cases given these and other antigiardial agents has raised concern and led to a search for other compounds. The aim of the present study was to compare the efficacy and safety, in the treatment of children infected with G. lamblia, of nitazoxanide, given at a dose of 7.5 mg/kg twice a day for 3 days, with those of tinidazole, given as a single dose of 50 mg/kg. Overall, 166 children, each proven to be infected with G. lamblia by the microscopical examination of a faecal sample, were included in the open and randomized trial, each being allocated to receive nitazoxanide or tinidazole. The parents of each treated child were asked to collect two faecal samples from the child between 5 and 10 days after the completion of treatment, for the parasitological follow-up. Only if no G. lamblia were found in both post-treatment samples from a child was that child considered cured. Among the 137 children who completed the study (74 given nitazoxanide and 63 given tinidazole), the frequency of parasitological cure following a single dose of tinidazole was significantly higher than that following six doses of nitazoxanide (90.5% v. 78.4%; P<0.05). Both treatment schedules were well accepted and well tolerated, with only mild, transient and self-limited side-effects reported. The commonest symptom on enrolment, diarrhoea, generally cleared 2-6 days after the initiation of treatment. Although apparently less efficacious than tinidazole, nitazoxanide remains a good candidate for the treatment of children with G. lamblia infection.